Andreas Delmelle | 1 May 2008 11:20
Picon
Favicon

Re: Fop Issue

On Apr 30, 2008, at 20:40, JimBME wrote:
>
> white-space-treatment="preserve" didn't seem to work but  white- 
> space="pre"
> did.

Just noticed: white-space-treatment probably doesn't work because of  
the additional white-space/wrap-option properties. If you remove  
those, it should also work.

'white-space' is a shorthand, that sets four properties
1) white-space-treatment
2) linefeed-treatment
3) white-space-collapse
4) wrap-option

see: http://www.w3.org/TR/xsl/#white-space

1) determines what should happen with spaces surrounding a linefeed  
or linebreak
2) determines what should happen with linefeed characters in the input
3) determines what should happen with series if space-characters
4) determines if implicit linebreaks should be generated

Cheers

Andreas
torsten.blix | 2 May 2008 10:49
Picon

Nested, multi-page tables and keep-with-previous

Hi,

I have a problem with the XSL-FO listed below - sorry for the lengthy example,
I tried to cut it down as much as possible (and hope that Outlook doesn't make
 too much of a mess of it).

Running FOP 0.95beta from the command line and producing a PDF output, I get no
warnings/errors and the expected text on the first page up until "LOOK HERE".
Then follows an empty second page and the end of the document, even though
there is some text left in the XSL-FO. I expected the text to be printed on the
2nd page.

Removing the attribute keep-with-previous.within-column='always' from the table
row below "LOOK HERE" makes the text on the 2nd page appear again.

FOP 0.94 behaves differently and does not hide any text, but breaks the table
apart a bit early imho (there seems to be plenty of place left at the bottom of
page 1).

The support for keep-with-previous should be partial, as per the compliance
page (http://xmlgraphics.apache.org/fop/compliance.html#fo-property-keep-with-previous).

The comments there read:
    * [0.20.5] works only in table rows
    * [0.94] works on all implemented FOs, except list- and table-related and
      inline-level FOs.
    * [0.95] works on all implemented FOs, except list- and inline-level FOs.
    * [0.94 and later] <integer> values are not supported.
I would expect my document not to work in 0.94 but to work in 0.95, as the usage
of keep-with-previous is table-related. Nevertheless, not showing parts of the
table is a bit unexpected.  Just ignoring the keep-with-previous attribute
(Continue reading)

Andreas Delmelle | 2 May 2008 11:08
Picon
Favicon

Re: Nested, multi-page tables and keep-with-previous

On May 2, 2008, at 10:49, <torsten.blix <at> sdm.de> <torsten.blix <at> sdm.de>  
wrote:

Hi

> I have a problem with the XSL-FO listed below - sorry for the  
> lengthy example,
> I tried to cut it down as much as possible (and hope that Outlook  
> doesn't make
>  too much of a mess of it).

No problem. It came through just fine, but maybe in the future it  
would be better if you just added it as an attachment.

> Running FOP 0.95beta from the command line and producing a PDF  
> output, I get no
> warnings/errors and the expected text on the first page up until  
> "LOOK HERE".
> Then follows an empty second page and the end of the document, even  
> though
> there is some text left in the XSL-FO. I expected the text to be  
> printed on the
> 2nd page.
>
> Removing the attribute keep-with-previous.within-column='always'  
> from the table
> row below "LOOK HERE" makes the text on the 2nd page appear again.

Definitely seems like buggy behavior in 0.95. Would you mind creating  
a Bugzilla entry for it, and add the FO as an attachment over there?
(Continue reading)

Vincent Hennebert | 2 May 2008 11:30
Favicon

Re: Nested, multi-page tables and keep-with-previous

Hi Torsten,

Torsten Blix wrote:
> Hi,
>
> I have a problem with the XSL-FO listed below - sorry for the lengthy example,
> I tried to cut it down as much as possible (and hope that Outlook doesn't make
>  too much of a mess of it).

FYI you can provide the FO file as an attachment. Actually this is the
recommended way because then you are sure your mail client won’t alter
it. Although in the present case it did well.

> Running FOP 0.95beta from the command line and producing a PDF output, I get no
> warnings/errors and the expected text on the first page up until "LOOK HERE".
> Then follows an empty second page and the end of the document, even though
> there is some text left in the XSL-FO. I expected the text to be printed on the
> 2nd page.

This is bug #44621 [1] (more precisely, comment #2). Assertions aren’t
enabled by default, otherwise you would have had an AssertionError. It
appears that FOP terminates normally but the generated PDF is obviously
wrong.

It is fixed in the head of the 0.95 branch, so your file should be
rendered fine with the final 0.95 version.

[1] https://issues.apache.org/bugzilla/show_bug.cgi?id=44621

> Removing the attribute keep-with-previous.within-column='always' from the table
(Continue reading)

Andreas Delmelle | 2 May 2008 11:40
Picon
Favicon

Re: Nested, multi-page tables and keep-with-previous

On May 2, 2008, at 11:30, Vincent Hennebert wrote:
>
> This is bug #44621 [1] (more precisely, comment #2). Assertions aren’t
> enabled by default, otherwise you would have had an AssertionError. It
> appears that FOP terminates normally but the generated PDF is  
> obviously
> wrong.
>
> It is fixed in the head of the 0.95 branch, so your file should be
> rendered fine with the final 0.95 version.

FWIW: It doesn't render fine with FOP Trunk...

If you remove the table-nesting, it also works as expected. If the  
outer table's first column is used solely for the purpose of spacing,  
then maybe removing the nesting and using start-indent could be a  
viable workaround.

Cheers

Andreas
torsten.blix | 2 May 2008 12:10
Picon

AW: Nested, multi-page tables and keep-with-previous

On May 2, 2008, at 11:41, Andreas Delmelle wrote:
> On May 2, 2008, at 11:30, Vincent Hennebert wrote:
> >
> > This is bug #44621 [1] (more precisely, comment #2). Assertions aren't 
> > enabled by default, otherwise you would have had an AssertionError. It 
> > appears that FOP terminates normally but the generated PDF is 
> > obviously wrong.
> >
> > It is fixed in the head of the 0.95 branch, so your file should be 
> > rendered fine with the final 0.95 version.
> 
> FWIW: It doesn't render fine with FOP Trunk...

As per Andreas' request I filed bug #44920.

> If you remove the table-nesting, it also works as expected. If the outer
> table's first column is used solely for the purpose of spacing, then maybe
> removing the nesting and using start-indent could be a viable workaround.

Yes, that would be a workaround. But as this effects a number of documents
which would have to be changed by hand (and are originally generated, thus not
really friendly on the eye) I would prefer a different solution ;)

Cheers,
Torsten
Vincent Hennebert | 2 May 2008 12:20
Favicon

Re: Nested, multi-page tables and keep-with-previous

Andreas Delmelle wrote:
> On May 2, 2008, at 11:30, Vincent Hennebert wrote:
>>
>> This is bug #44621 [1] (more precisely, comment #2). Assertions aren’t
>> enabled by default, otherwise you would have had an AssertionError. It
>> appears that FOP terminates normally but the generated PDF is obviously
>> wrong.
>>
>> It is fixed in the head of the 0.95 branch, so your file should be
>> rendered fine with the final 0.95 version.
>
> FWIW: It doesn't render fine with FOP Trunk...

This is because the fix was made in the 0.95 branch and hasn’t been
merged back to the Trunk yet. I tried to render the file with the ‘head’
version of the 0.95 branch and the result was ok.

> If you remove the table-nesting, it also works as expected. If the outer
> table's first column is used solely for the purpose of spacing, then
> maybe removing the nesting and using start-indent could be a viable
> workaround.
>
>
> Cheers
>
> Andreas

Vincent

--

-- 
(Continue reading)

Andreas Delmelle | 2 May 2008 12:26
Picon
Favicon

Re: Nested, multi-page tables and keep-with-previous

On May 2, 2008, at 12:20, Vincent Hennebert wrote:
> Andreas Delmelle wrote:
>>
>> FWIW: It doesn't render fine with FOP Trunk...
>
> This is because the fix was made in the 0.95 branch and hasn’t been
> merged back to the Trunk yet. I tried to render the file with the  
> ‘head’
> version of the 0.95 branch and the result was ok.

OK, my bad... Seems like Torsten is in luck. :-)

Cheers

Andreas
Vangelis Karageorgos | 2 May 2008 21:39
Picon

Page bottom padding

Hi everybody,

there's a question we've been confronted with during our development process using FOP.
Is page bottom padding possible to materialize in FOP, while constructing a layout with several columns and flow from one onto another?

Thanks in advance,

Philipp Wagner | 5 May 2008 00:14

tables and keeps with 0.95beta

Hi,

I tried FOP 0.95beta together with a table with 
keep-together.within-page="always" and keep-with-previous="always" 
attributes, which didn't work in 0.94 at all. While the page breaks are 
almost correct now, there is a problem with the table headers. I 
uploaded a sample document at
http://philipp-wagner.com/temp/20080504-fop/document.fo
and the PDF output at
http://philipp-wagner.com/temp/20080504-fop/document.pdf

If you look on page two you'll see a horizontal line between the first 
and the second row, which shouldn't be there.

Apart from that, the keep-with-previous="always" attribute doesn't work, 
as you can see on the top of the second page, too. The first row on the 
second page has a keep-with-previous attribute, which should keep it 
together with the row before, which is not the case here.

Are this known limitations of FOP or just a small bug? Or are there any 
workarounds?

Philipp

Gmane